Howdy.

Another user idea to be sequestered to "maybe in GC3" (if it even gets noticed).

Approval should react differently to tax changes. The problems with the artificial population in GC2 are: a) they react with total naivety to changes in tax; with sudden and precipitous changes they completely forget what the tax rate was like a week ago; and their happiness is absolutely, linearly proportional to the tax rate.

Now, I'm not a psychologist (nor an economist), but I think a real population would not react in this way to taxes. If the President of Zimbabwe suddenly drops taxes to next-to-nothing just before the elections, people would not suddenly think he's turned into a saint and vote him in again. If anything, everyone would realise it's a shameless ploy and try to vote him out even more. Likewise, how I accept taxes are they are now is a balance between the raw percentage and factors like what the rate was like last year or ten years ago, and what people next door are paying, etc.

1. A tax rate that remains stable over long periods should be reason for an approval bonus itself. Conversely, a constantly fluctuating rate would make people grumble because they can't do their budgets properly anymore.

2. Decrements to tax rate should react thus:

a) Small changes (1-4%) would make the populace grateful of the government for the "token" reduction (since it would not mean immense financial benefits to the people; however, everyone likes to ease off some pressure). Thus the benefit to approval should be slightly less than the linear model, and the bonus would not last long before tending back toward the pre-change approval.

Medium changes (5-8%) would give an average reduction benefit and would not cause a drift back to the "norm".

Larger changes (9-20%) would not only make people like the government, but they will also begin realising that they're getting much richer. Thus, an immediate morale improvement would be seen, and there would be a longer term compounded bonus up to a certain approval figure (what I mean is, the approval bonus from large decrements would continue to increase over X turns, up to a certain point -- it would be easy to make this end-point the appropriate figure in the current linear model, though it need not be).

c) Very large decrements (>20%) will not stretch the "upward-drift" much further, but will show bigger bonuses.

3. Increments should work much the same:

a) Small increments will make 'em unhappy but the approval will eventually stabilise not far from the pre-change figure as people get used the tax rate.

Medium increments will penalise morale but approval will not "drift".

c) Large increments, again, will show a sustained penalty over a period of time until it stabilises again.

4. Lastly, longer trends must have an influence on all the above possibilities. The week-before-elections reduction must take into account the average tax rate of the previous year, so that people will not be so easily bought over and approval will be resistant to bonuses for a while. Conversely, a sustained low tax rate will make approval reactions to increments much more tolerant.

In summary, the idea is that approval will still show a immediate (albeit smaller) reaction to changes in tax, but will also have a longer-term "drift" effect. Not only will this prevent exploits, it will also reward thoughtful taxation and punish negligence, in addition to just being more realistic.

Comments are welcome.
J.